Flight Status Predictor

Welcome to Flight Status Predictor, an application that predicts if a flight may be cancelled, delayed, or on time. It does so using decision trees trained on 6 million tuples containing flight information.

Prerequisites

Python 3.10.12 or greater.
Jupyter Lab

Instructions

Install Models and Dictionaries

The jupyer notebook should contain the following cell already, which should install all dependencies:

!pip install pandas
!pip install scikit-learn
!pip install matplotlib
!pip install imbalanced-learn
!pip install joblibspark

Run the final notebook provided. The notebook will create the following pickles:

Airline_unique.pickle
Dest_unique.pickle
DestCityName_unique.pickle
DestStateName_unique.pickle
flight_predictor_model.pickle (model)
Month_unique.pickle
Origin_unique.pickle
OriginCityName_unique.pickle
OriginStateName_unique.pickle

Save these on a "/pickles/" folder on the same level as the src directory (as in the repo).

Install dependencies via

pip install sklearn
pip install pickle
pip install tkinter
pip install ttkbootstrap

Run

python3 main.py